Main

Animal studies indicate that experience-related changes may occur in mammalian brain structures, but so far there has been no evidence of comparable alterations in the human brain3,4,5. To investigate this possibility, we divided a homogeneous group of volunteers (21 female, 3 male; mean age, 22 yr ± 1.6 s.d.), who were matched for sex and age, into two groups, designated as jugglers and non-jugglers. Both groups were inexperienced in juggling at the time of their first brain scan.

Subjects in the juggler group were given 3 months to learn a classic three-ball cascade juggling routine. A second brain scan was performed when they had become skilled performers (that is, when they could sustain juggling for at least 60 seconds). A third scan was carried out 3 months later; during the intervening period, none of the jugglers practised or attempted to extend their skills — for example, by learning a four-ball or a reverse cascade. In fact, most subjects were no longer fluent in three-ball cascade juggling by the time of the third scan.

We used voxel-based morphometry, a sophisticated objective whole-brain technique, to investigate subtle, region-specific changes in grey and white matter by averaging results across the volunteers. This method is based on high-resolution, three-dimensional magnetic-resonance imaging, registered in a common stereotactic space, and is designed to find significant regional differences by applying voxel-wise statistics in the context of gaussian random fields6,7.

Group comparison at the beginning (the baseline) showed no significant regional differences in grey matter between jugglers and non-jugglers. In the longitudinal analysis, the juggler group demonstrated a significant (44 d.f., P<0.05) transient bilateral expansion in grey matter in the mid-temporal area (hMT/V5) and in the left posterior intraparietal sulcus between the first and the second scans. This expansion decreased in the third scan (Fig. 1). We found a close relationship in these regions between the transient structural grey-matter changes and the juggling performance. These findings were specific to the training stimulus, as the non-jugglers showed no change in grey matter over the same period.

Figure 1: Transient changes in brain structure induced while learning to juggle.
figure 1

a–c, Statistical parametric maps showing the areas with transient structural changes in grey matter for the jugglers group compared with non-juggler controls. a, Sagittal view; b, coronal view; c, axial view. The increase in grey matter is shown superimposed on a normalized T1 image. The left side (L) of the brain is indicated. A significant expansion in grey matter was found between the first and second scans in the mid-temporal area (hMT/V5) bilaterally (left: x, −43; y, −75; z, −2, with Z = 4.70; right: x, 33; y, −82; z, −4, with Z = 4.09) and in the left posterior intraparietal sulcus (x, −40; y, −66; z, 43 with Z = 4.57), which had decreased by the time of the third scan. Colour scale indicates Z scores, which correlate with the significance of the change. d, Relative grey-matter change in the peak voxel in the left hMT for all jugglers over the three time points. The box plot shows the standard deviation, range and the mean for each time point.

Full size image

Our results contradict the traditionally held view that the anatomical structure of the adult human brain does not alter, except for changes in morphology caused by ageing or pathological conditions. Our findings indicate that learning-induced cortical plasticity is also reflected at a structural level.

As all of our volunteers have normal fine-motor skills, we conclude that juggling, and consequently the perception and spatial anticipation of moving objects, is a stronger stimulus for structural plasticity in the visual areas (used for the retention of visual-motion information8,9) than in the motor areas (involved in the planning and execution of coordinate motion — that is, the supplementary motor area and/or the motor cortex, cerebellum and basal ganglia).

Although the observed transient increase in grey matter takes place in specific motion-selective areas, the microscopic changes underlying these dynamic structural alterations are unclear. Macroscopic alterations may be based on changes at the level of synaptic bulk and neurites, or they might include increased cell genesis, for example, of glial or even neuronal cells4. Imaging results need to be compared with histological data for identification of the structural basis at the microscopic level of temporary, training-dependent structural changes in our brains.